butjer1010 wrote on 2026-01-20, 23:04:

My only advice for You is - Listen to mkarcher :) This guy is a genius, he did manage to help me repair EGA graphics card, which […]
Show full quote

My only advice for You is - Listen to mkarcher 😀
This guy is a genius, he did manage to help me repair EGA graphics card, which i thought it is for trash only! And i'm not an electrician, i'm just a guy who knows how to solder, and that's it! He did everything!
Try everything he will ask here, and there is a big chance to get this baby work again 😀
I only have one question - without RAM sticks, is there any sound on speaker?

Yep. mkarcher helped me also on another 286 board and we got it figured out. I'm thankful to him as well all others who are trying to help with suggestions 👍

Without RAM, no difference.. POST card is still not showing any codes and no beeper sounds (except for that little click I mentioned when turning on/off/resetting).

mkarcher wrote on 2026-01-21, 00:27:

It's nice you have a scope. That makes troubleshooting easier, but this picture means that something is severely off or you don' […]
Show full quote

wbahnassi wrote on 2026-01-20, 22:55:

Yes, I'm a noob with the oscilloscope, but I think I managed to capture the /ADS signal on reset and it indeed behaves as you described. For this, I switched the oscilloscope's mode to Single and trigger on a rising edge. Upon reset, it immediately captured the profile you can see in the photo and paused. /ADS goes low for 150ms, then goes back to 5V and stays there.

It's nice you have a scope. That makes troubleshooting easier, but this picture means that something is severely off or you don't really know what you are doing with the scope. i suspect the issue is the latter. My primary issues with that image are:

• You set triggering to the rising edge (you say so, and the screenshot icon confirms it), yet I see a falling edge at the center of the screen. The orange "T" inside an arrow at the top of the grid confirms that the center of the screen displays what happened at the trigger time.
• You see a pulse of 150 milliseconds (as I understand the screenshot, you are reading it correctly, but /ADS should be low for an individual clock cycle of the processor to start a bus cycle, which is 40 nanoseconds.

I trust the scope to trigger on a rising edge. So there is a rising edge where you don't see any. The primary reason is: The time base (horizontal resolution) is way off a setting that would give good measurements. The signal on /ADS consists of sub-microsecond pulses, yet you configured the scope to display 100.000 microseconds per grid line. That's the "M: 100ms" in the top. You should be able to "zoom into" it by setting the time base to 100 nanoseconds instead of 100 milliseconds per grid line, and retry the measurement. I expect you will see a lot of pulses (at least that's what is expected). The issue you are facing here is a kind of aliasing caused by sampling way too little points to properly resolve what's going on. It's an easy trap to fall into, and if the general shape (like a pulse here) makes sense at first, hard to notice that you are not measuring what you think you are measuring. The seemingly wrong triggering behaviour might give a clue if you have some scope experience.

So, please redo that measurement. Single shot and vertical settings are fine, or at least may be fine. The scope screenshot doesn't show whether you are probing in 1x or 10x mode. For signals that exceed a couple of MHz (that is, pulses shorter than 300 nanoseconds), probing with your standard passive scope probes at 1x gives distorted results. It seems the scopes shipped with your scope are switchable. You only get the 70 MHz bandwidth (which we need) at the 10x setting. The 1x setting is typically limited to around 6 to 10 MHz. If you configure the probe into 10x mode, you either need to configure the scope accordingly (this is possible in all digital scopes I know), or you need to be aware that 5V will be displayed as 0.5V instead.

So, assuming you see some individual short pulses on /ADS, it seems like the CPU tries to access the BIOS to get started. As you have a scope, the next thing you should check is whether the BIOS is adressed. The BIOS chip has chip enable on pin 20 and output enable on pin 22. Both of these pins need to be low to get data from the BIOS. Please check whether this is the case.

Yeah I'm positive it's me being the problem here. I'm still learning to use the scope, and this is the first time I use it outside of Auto mode.

My probe is set to 10x, and the scope is also set to 10x to match the probe. I have readjusted the horizontal time scale to 100ns.. now the probe is not triggering at all upon resetting the motherboard. Tried several times and I get no triggers at all. I had the trigger on 2.5V then again on 4.5V... still no triggers.. line is always on 5V.

Now, on the BIOS chip I see both pins 20 and 22 are high, not low.. They don't budge whether I reset or not.. just always high. Seems like I have a big problem here. What drives those pins?

Traced the BIOS pins 20 and 22 to two different pins on the main chipset... seems like I'm back to the chipset again... why is it keeping those high?

wbahnassi wrote on 2026-01-21, 02:25:

My probe is set to 10x, and the scope is also set to 10x to match the probe. I have readjusted the horizontal time scale to 100ns.. now the probe is not triggering at all upon resetting the motherboard. Tried several times and I get no triggers at all. I had the trigger on 2.5V then again on 4.5V... still no triggers.. line is always on 5V.

This would be back to square one, then. While last time, you observed activity on /ADS, you no longer do, so the system is in a worse state than previously. Just to confirm that something changed, you might want to retry the measurement at a 100ms timescale, but I would be extremely surprised if you get a trigger at 100ms, but no trigger at 100ns.

"Back to square one" means, please test whether RESET (active high while the reset button is pressed), CLK and Vcc are still OK, also re-do the HOLD/AHOLD/BOFF# check.

wbahnassi wrote on 2026-01-21, 02:25:

Now, on the BIOS chip I see both pins 20 and 22 are high, not low.. They don't budge whether I reset or not.. just always high. Seems like I have a big problem here. What drives those pins?

That's perfectly OK while the bus is idle. As long as you don't get the 486 processor to read the BIOS (which will require pulses on /ADS), probing the BIOS chip is pointless. Seeing both pins at solid 5V likely means that the chipset is driving the idle state correctly.

Check for bad decoupling caps.
Caps fail short.
If as you say, a line is never going high, that would be the simplest cause and solution.

wbahnassi wrote on 2026-01-20, 22:55:

For the RAM slots, the board has two sets: 2 72-pin SIMMs and 4 30-pin SIMMs. I was using 72-pin memory sticks.. but I just tried a set of four 30-pin sticks and it didn't make a difference. Usually bad memory would cause the board to make continuous beeps.. but here it's just dead silent.

I had a board that was silent (no response or beeps) until I cleaned the RAM slots with isopropyl alcohol & a toothbrush. Reseating the sticks did nothing.

wbahnassi wrote on 2026-01-21, 02:25:

Now, on the BIOS chip I see both pins 20 and 22 are high, not low.. They don't budge whether I reset or not.. just always high. Seems like I have a big problem here. What drives those pins?

Have you taken out & reseated the BIOS chip? Every board I ever owned needed that.

myne wrote on 2026-01-21, 07:13:

Check for bad decoupling caps.
Caps fail short.
If as you say, a line is never going high, that would be the simplest cause and solution.

Yes I wanted to do the check last evening and I just got to it. I checked every cap thoroughly.. they are all not reading any shorts... except... the very last one I checked.. this big blue dude:

I'll remove it and try without it for a short period.

mkarcher wrote on 2026-01-21, 06:56:

This would be back to square one, then. While last time, you observed activity on /ADS, you no longer do, so the system is in a […]
Show full quote

wbahnassi wrote on 2026-01-21, 02:25:

My probe is set to 10x, and the scope is also set to 10x to match the probe. I have readjusted the horizontal time scale to 100ns.. now the probe is not triggering at all upon resetting the motherboard. Tried several times and I get no triggers at all. I had the trigger on 2.5V then again on 4.5V... still no triggers.. line is always on 5V.

This would be back to square one, then. While last time, you observed activity on /ADS, you no longer do, so the system is in a worse state than previously. Just to confirm that something changed, you might want to retry the measurement at a 100ms timescale, but I would be extremely surprised if you get a trigger at 100ms, but no trigger at 100ns.

"Back to square one" means, please test whether RESET (active high while the reset button is pressed), CLK and Vcc are still OK, also re-do the HOLD/AHOLD/BOFF# check.

wbahnassi wrote on 2026-01-21, 02:25:

Now, on the BIOS chip I see both pins 20 and 22 are high, not low.. They don't budge whether I reset or not.. just always high. Seems like I have a big problem here. What drives those pins?

That's perfectly OK while the bus is idle. As long as you don't get the 486 processor to read the BIOS (which will require pulses on /ADS), probing the BIOS chip is pointless. Seeing both pins at solid 5V likely means that the chipset is driving the idle state correctly.

I just did a recheck.. no trigger on /ADS. HOLD and AHOLD are low, BOFF is high. RESET responds correctly to reset switch. Clock is good. VCC is good.

Let me check on that cap and come back to you..

In the mean time, just to confirm.. ADS pin I'm probing is at S 17 in this pinout diagram. Correct? There is also EADS on B 17 but I'm not probing that.

Ok removed the cap, but the short on its location on the board is still there. The cap itself is not shorted outside the board. The short is a perfect short (0 Ohm resistance). Still same dead behavior on the board.
I'll keep the cap out for a while until I find where this short is coming from.

Ok, the short goes away when I remove the jumpers on U50.

These jumpers control the CPU voltage according to the manual. I had them set to close 3-6 and 4-5.

I'm guessing then this short is "expected behavior", and there's nothing wrong here?

I'm attaching the piece from the manual that talks about this jumpers block.

Just for the fun of it I powered the board with the jumpers removed, but I guess that's even more reason for the CPU to refuse to work as it is severely under-volted.

wbahnassi wrote on 2026-01-21, 13:52:

I just did a recheck.. no trigger on /ADS. HOLD and AHOLD are low, BOFF is high. RESET responds correctly to reset switch. Clock is good. VCC is good.

That's unexpected. The logic inside the 486 processor is supposed to trigger a code fetch from Address FFFFFFF0 within approximately 1 millions clock cycles after RESET goes low. Unless the processor is kept off the bus, the code fetch is supposed to get to the board.

wbahnassi wrote on 2026-01-21, 13:52:

In the mean time, just to confirm.. ADS pin I'm probing is at S 17 in this pinout diagram. Correct? There is also EADS on B 17 but I'm not probing that.

That's completely correct. /ADS is driven by the current owner of the 486 front side bus (VL master, 486 processor) whenever a cycle starts. On the other hand, /EADS is driven by the mainboard during address hold to start a cache invalidation cycle inside the 486 processor, for example when ISA DMA writes to memory.

I wonder what's the best way to continue. The processor on the picture in the first post is a "classic" 486 processor without system management mode, so it doesn't have a SRESET line ("soft reset", which will reset the CPU logic but keep L1 intact). So the tests you did (Vcc, CLOCK, HOLD, AHOLD, BOFF) should suffice for the processor to generate /ADS. Something very basic thus must be wrong, like a broken processor. I've not yet seen a dead 486-33, so I don't think "broken processor" is likely. The board has a ZIF socket, and I had issues with insufficiently inserted processors when working on 486 or Pentium boards multiple times, so re-inserting the processor might be a good idea. Also, please note that the processor socket on your board is a socket 3, with 237 pins, not the original PGA168 socket which you showed the pinout for. Socket 3 has an extra row of pins on the outside, so /ADS is not located at the extreme edge of that socket, but one row/column further inwards. Make sure you are probing the correct pin. The location of /ADS makes it very unlikely that you accidentally mixed up rows and columns, so i don't think this is a likely cause for your issue.

Anyway, something clearly changed between the time when you made the oscilloscope picture with /ADS being displayed low for 150ms and now, which is a change to the worse. I suggest you find the reason for that change first. If you re-inserted the processor inbetween, a insufficiently inserted processor is a quite probable cause for the symptoms you are seeing, so re-inserting the processor again and making sure you probe the correct pin /ADS would be the first step I suggest.

That cap you remove is leaking. I can see liquid .. chack the traces underneath